The invention relates to a phase-detection autofocus system using a single two-dimensional (2D) image sensor as an autofocus sensor. This allows for a number of novel functionalities such as calibration by adaptive selection of pixels for the autofocus collection zone, dynamically optimization of the collection zone for various scene and lightning conditions, as well as the use of 2D phase signals in the cross-correlation. This freedom in selecting which pixels to use gives rise to an equivalent freedom for the focus points presented to the user, which can be created, moved and scaled according to the conditions and the scene.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method for generating a focus measurement signal for a taking lens of a camera, in a phase-detection autofocus system, the method comprising: focusing at least one pair of luminous bundles from the lens onto a corresponding pair of illuminated regions on an autofocus sensor, wherein the autofocus sensor is a single high-resolution, two-dimensional, CMOS image sensor separate from a taking sensor of the camera, the CMOS image sensor comprising a single continuous, two-dimensional array of light sensitive pixels, each pixel being associated with an on-chip, light-insensitive storage element; determining values from the illuminated regions; performing a correlation analysis between values of pixels in at least two pairs of the illuminated regions to generate a focus measurement signal, wherein a result of the correlation analysis yields a correlation result for each pair of the at least two pairs of illuminated regions; and comparing the correlation results from the at least two pairs of illuminated regions to identify pixels to use in generating the focus measurement signal.
2. The method according to claim 1 , further comprising selecting the pixels used in the correlation analysis dynamically and/or adaptively from the illuminated regions.
3. The method according to claim 1 , wherein performing the correlation analysis comprises performing a calibration by at least: placing a target in a field of view of the phase-detection autofocus system, bringing the target into focus, and determining a displacement between depictions of the target in the illuminated regions.
4. The method according to claim 3 , wherein a direction between the illuminated regions is defined as a first direction and a second direction is defined as being orthogonal to the first direction and in a plane of the image sensor, the method further comprising selecting sets of pixels to be used in the correlation analysis to have a displacement equal to a component of the determined displacement in the second direction.
5. The method according to claim 2 , wherein selecting the pixels comprises: selecting a first set of pixels and performing the correlation analysis, and if a result of the correlation analysis is poor, selecting a second set of pixels and performing the correlation analysis.
6. The method according to claim 2 , wherein selecting the pixels comprises: for each of a plurality of correlations, performing a correlation analysis between values of a different set of a plurality of sets of pixels in the regions, scoring the plurality of correlations, and selecting one or more sets of pixels to be used in generating the focus measurement signal.
7. The method according to claim 1 , further comprising: increasing an effective pixel size by combining adjacent pixels within each illuminated region to accommodate for light conditions, wherein the combining generates values of combined pixels; and performing the correlation analysis between the values of the combined pixels in the regions.
8. The method according to claim 1 , wherein the correlation is a two-dimensional correlation to determine an image pattern present in the pair of illuminated regions and a displacement between their positions within their respective regions.
9. A phase-detection autofocus system for detecting a focus of a taking lens of a camera, the system comprising: a single high-resolution, two-dimensional CMOS image sensor separate from a taking sensor of the camera, the image sensor comprising a single continuous, two-dimensional array of light sensitive pixels, each pixel being associated with an on-chip, light-insensitive storage element; optical elements for focusing at least one pair of luminous bundles from the lens onto a corresponding pair of illuminated regions on the image sensor; and an evaluation unit configured for: performing a correlation analysis between values of pixels in at least two pairs of the illuminated regions to generate a focus measurement signal, wherein a result of the correlation analysis yields a correlation result for each pair of the at least two pairs of illuminated regions, and comparing the correlation results from the at least two pairs of illuminated regions to identify pixels to use in generating the focus measurement signal.
10. The phase-detection autofocus system according to claim 9 , further comprising: a CMOS readout circuit configured to determine values of pixels from the illuminated regions from the sensor; and wherein the evaluation unit is configured for dynamically and/or adaptively selecting the at least some of the pixels from the illuminated regions to be used in the correlation analysis.
11. The method according to claim 1 , further comprising: determining a first correlation result for a first pair of illuminated regions; determining that the first correlation result does not satisfy correlation result criteria; in response to determining that the first correlation result does not satisfy the correlation result criteria, determining a second correlation result for a second pair of illuminated regions; determining that the second correlation result does satisfy the correlation result criteria; and selecting pixels from the second pair of illuminated regions to use in generating the focus measurement signal.
12. The method according to claim 1 , further comprising: determining a first weighting for a first pair of illuminated regions based on a first correlation result of the first pair of illuminated regions; determining a second weighting for a second pair of illuminated regions based on a second correlation result of the second pair of illuminated regions; and using the first and second weightings to generate the focus measurement signal based on pixels from both of the first and second pairs of illuminated regions.
13. The method according to claim 1 , further comprising: determining a orientation of a line of pixels in each of a first pair of illuminated regions; determining a second orientation of a line of pixels in each of a second pair of illuminated regions, wherein the second orientation is different from the first orientation; and generating the focus measurement signal based on pixels from both of the first and second pairs of illuminated regions.
14. The method according to claim 1 , further comprising combining values from adjacent pixels within each illuminated region to compensate for low lighting conditions.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
May 27, 2016
December 18, 2018
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.